Article

pubs.acs.org/JAFC

Influence of Oak Maturation Regimen on Composition, Sensory

Properties, Quality, and Consumer Acceptability of Cabernet
Sauvignon Wines
Anna M. Crump, Trent E. Johnson, Kerry L. Wilkinson,* and Susan E. P. Bastian
School of Agriculture, Food and Wine, The University of Adelaide, PMB 1, Glen Osmond, SA 5064, Australia

ABSTRACT: Oak barrels have long been the preferred method for oak maturation of wine, but barrels contribute significantly
to production costs, so alternate oak maturation regimens have been introduced, particularly for wines at lower price points. To
date, few studies have investigated consumers’ acceptance of wines made using non-traditional oak treatments. In this study, two
Cabernet Sauvignon wines were aged using traditional (i.e., barrel) and/or alternative (i.e., stainless steel or plastic tanks and vats,
with oak wood added) maturation regimens. Chemical and sensory analyses were subsequently performed to determine the
influence on wine composition and sensory properties, that is, the presence of key oak-derived volatile compounds and
perceptible oak aromas and flavor. The quality of a subset of wines was rated by a panel of 10 wine experts using a 20-point
scoring system, with all wines considered technically sound. Consumer acceptance of wines was also determined. Hedonic ratings
ranged from 5.7 to 5.9 (on a 9-point scale), indicating there was no significant difference in consumers’ overall liking of each
wine. However, segmentation based on individual liking scores identified three distinct clusters comprising consumers with
considerably different wine preferences. These results justify wine producers’ use of alternative oak maturation regimens to
achieve wine styles that appeal to different segments of their target market.
KEYWORDS: oak, wine, descriptive analysis, preference, wine consumers

■ INTRODUCTION
Consumer research is often used to gain insight into the market
acceptance, preference, and perception of different foods and
beverages, in particular, to establish the relative influence of
extrinsic (e.g., packaging, labeling, and branding) and/or
intrinsic (e.g., sensory attributes) characteristics on consumer
liking and purchase intent.1 For example, consumer studies
have been undertaken to determine the influence of different
bottle closures (i.e., natural cork, synthetic cork, and screw cap)
on consumers’ perceptions of wine quality and price.2
Consumers were unable to differentiate wines bottled under
different closures during blind tastings, but consumer liking and
quality ratings of wines bottled under screw cap decreased
significantly when the closure type was disclosed.2 This study
demonstrates the importance of understanding the implications
of production decisions, in this case packaging, on consumer
acceptability.
Market segmentation can also be used to differentiate
consumer groups within a given population. For example,
previous studies have segmented wine consumers according to
their lifestyles or their wine experience, knowledge, or
involvement.3−7 This enables the inherent heterogeneity of a
population to be taken into account, and in some cases, for
product development and/or marketing strategies to be
targeted toward specific groups within a market. The use of
oak in winemaking provides another good example of the
potential application of consumer research.
Oak barrels have long been the preferred method for oak
maturation of wine. During barrel maturation, volatile
compounds are extracted into the wine and impart desirable
oak aromas and flavors,8,9 whereas the controlled oxidation and
modification of tannins and anthocyanins are considered to
improve the intensity, stability, and structure of wines.10,11 As a
consequence, barrel-aged wines are generally considered to be
of higher quality and therefore attract higher prices.12 However,
because the use of barrels also contributes to significantly
higher production costs, wine producers have increased their
use of alternative oak maturation regimens. Several studies have
reported the evolution of oak-derived volatiles and sensory
attributes in wine following treatment with oak chips,13,14 in
some cases with micro-oxygenation,15,16 so as to achieve
maturation conditions that more closely replicate traditional
barrel aging.
To date, few studies have investigated consumers’ acceptance
of wines made using non-traditional oak treatments. A recent
study compared consumer preferences for wines aged either in
oak barrels or with oak chips and reported considerable
disparity in consumers’ wine preferences.17 Because consumers
did not significantly reject wines made with oak chips, the
authors concluded that markets exist for wines made using both
oak maturation regimens. The role of oak in winemaking is not
well understood by most wine consumers, but knowledgeable
consumers do appreciate traditional barrel maturation and are
willing to pay a premium for barrel-aged wines, whereas less
knowledgeable consumers are accepting of the use of oak
chips.18 A greater understanding of consumer liking for wines
made using oak alternatives would enable the wine industry to
better ascertain their acceptability in the market. This study
therefore aimed to investigate the influence of a range of
Received: September 21, 2014
Revised: January 13, 2015
Accepted: January 13, 2015
Published: January 13, 2015

© 2015 American Chemical Society 1593 DOI: 10.1021/jf5044025

J. Agric. Food Chem. 2015, 63, 1593−1600
Journal of Agricultural and Food Chemistry Article

Table 1. Regimens Used for Oak Maturation of 2009 and 2011 Cabernet Sauvignon Wines
wine sample code maturation vessel volume (L) oak maturation regimen
trial 1 2009 Padthaway SSVAT stainless steel vat 900 2 g/L toasted French oak chips
Cabernet Sauvignon P1VAT stainless steel vat with two 4.5 mm 900 2 g/L toasted French oak chips
HDPE panels
P2VATa stainless steel vat with two 1.5 mm 900 2 g/L toasted French oak chips
HDPE panels
FOVAT stainless steel vat with two French 900 1.6 m2 French oak panels
oak panels

trial 2 2011 Eden Valley barrela barrel 225−300b French oak (22% new, 16% 1-year-old, and 62% 2-year-
Cabernet Sauvignon old), medium toast
SS + OAKa stainless steel 4550 French oak staves (1.4 m2/kL), medium toast
SS − OAK stainless steel 4550 no oak treatment
P1 + OAKa plastic tank (with 12 g/L/year O2 1000 French oak staves (1.4 m2/kL), medium toast
ingress)
P1 − OAKa plastic tank (with 12 g/L/year O2 1000 no oak treatment
ingress)
P2 + OAK plastic tank (with 4 g/L/year O2 1000 French oak staves (1.4 m2/kL), medium toast
ingress)
P2 − OAK plastic tank (with 4 g/L/year O2 1000 no oak treatment
ingress)
a
Wines retained for consumer acceptability and expert panel testing. bBlend of wines aged in 4 × 300 L hogsheads and 26 × 225 L barriques.

traditional and alternative oak maturation regimens on the
composition, sensory properties, quality, and consumer accept-
ability of wine. The knowledge gained from this study can be
used by wine producers to both inform and justify their use of
oak alternatives, in particular for the production of wines at
certain price points and/or targeted toward specific segments of
the wine market.
an end point of pH 8.2), volatile acidity (VA, as acetic acid equivalents
to an end point of pH 8.2), free and total SO2 content (by the
aspiration method), wine color density, and total phenolics, according
to methods described previously.19 Glucose and fructose concen-
trations were determined enzymatically (Boehringer-Mannheim/R-
BioPharm, Darmstadt, Germany). Ethanol content (as % alcohol by
volume) was measured with an alcolyzer (Anton Paar, Graz, Austria).
Oak volatiles were quantified by gas chromatography−mass

■
MATERIALS AND METHODS
Oak Maturation Trials. Two South Australian Cabernet
Sauvignon wines were sourced from industry-based trials, involving a
range of different oak maturation regimens (Table 1). Trial 1 involved
a 12 month maturation of a 2009 Padthaway Cabernet Sauvignon in
four types of 900 L Stakvats (Ausvat Pty. Ltd., Willunga, Australia),
being (i) stainless steel, hereafter referred to as “SSVAT”; (ii) stainless
steel with two 4.5 mm high-density polyethylene (HDPE) panels,
hereafter referred to as “P1VAT”; (iii) stainless steel with two 1.5 mm
HDPE panels, hereafter referred to as “P2VAT”; and (iv) stainless
steel with two French oak wood panels, hereafter referred to as
“FOVAT”. French oak chips (medium toast) sourced from the same
batch of oak wood were inserted in each of the stainless steel and
plastic Stakvats (SSVAT, P1VAT, P2VAT), but not the oak-sided
Stakvat (FOVAT), at a rate of 2 g/L. Trial 2 involved a 4 month
maturation of a 2011 Eden Valley Cabernet Sauvignon in four different
vessels, being (i) 1000 L plastic tanks with oxygen ingress of 17 mg
O2/L/year, hereafter referred to as “P1”; (ii) 1000 L plastic tanks with
oxygen ingress of 12.6 mg O2/L/year, hereafter referred to as “P2”;
(iii) 4550 L stainless steel tanks, hereafter referred to as “SS”; and (iv)
a combination of new, 1- and 2-year-old French oak barrels (medium
plus toast), which were blended after maturation, hereafter referred to
as “barrel”. Maturation in plastic and stainless steel tanks comprised
treatments with and without the addition of French oak staves
(medium toast), sourced from the same batch of oak wood (hereafter
designated “+ OAK” and “− OAK”, respectively). Staves were added
to achieve an area-to-volume ratio of 1.4 m2/kL.
Following completion of each maturation trial, wines were bottled
(using 750 mL bottles fitted with metal screw-cap closures for trial 1
and 375 mL bottles fitted with plastic screw-cap closures for trial 2)
and stored at 15 °C prior to sensory and chemical analysis. Trial 1
wines were bottle aged for 3 years prior to analysis, whereas trial 2
wines were analyzed approximately 4 months after bottling.
Chemical Analysis of Wine. Wines were analyzed (in duplicate)
to determine pH, titratable acidity (TA, as tartaric acid equivalents to
spectrometry (GC-MS) using stable isotope dilution analysis (SIDA)
methods reported previously.20−22 These publications describe the
preparation of deuterated internal standards, method validation, and
instrumental operating conditions. Analysis of oak volatiles was
performed by the Australian Wine Research Institute (AWRI)
Commercial Services Laboratory using an Agilent 6890 gas chromato-
graph coupled to a 5973 mass selective detector.
Sensory Analysis of Wine. Wines were equilibrated at 21 °C for
24 h prior to each sensory experiment. Wines were assigned a random
three-digit code for descriptive analysis (DA), expert and consumer
sensory assessments. Wine aliquots (30 mL) were presented in
covered Viticole XL5 (ISO standard) 215 mL stemmed wine tasting
glasses, using a randomized presentation order.
Descriptive Analysis. DA was undertaken using a panel comprising
10 University of Adelaide staff and students (seven females, three
males). Panelists were recruited on the basis of their availability and
previous participation on DA panels involving analysis of red wine.
Prior to formal assessment, panelists underwent 9 h of training
(comprising 6 × 1.5 h sessions held over 6 weeks) involving detection,
identification, evaluation, and intensity rating of red wine aroma and
palate (i.e., flavor, taste and mouthfeel) attributes. During the training
sessions, the panel evaluated each wine at least twice, to generate
appropriate aroma and palate terms and to gain familiarity in
recognizing and scoring their intensity, according to previously
published protocols.23 The training sessions also included three
practice evaluation sessions conducted in sensory booths, under the
conditions used during formal assessment. Data were evaluated after
each practice evaluation to assess panelist-by-sample interactions, to
determine when formal assessment should commence. The panel
agreed upon 15 aroma attributes and 16 palate attributes; when
reference standards were used, their preparation is shown in Table 2.
Because panelists were unable to perceive any differences in wine
color, appearance was not evaluated. Three formal assessment sessions
were held, with all 11 wines presented in each session, such that all
wines were assessed in triplicate. Panelists assessed each wine in an
isolated tasting booth at 22−23 °C under red lights, with a light
temperature of 6500 K, and rated the intensity of each sensory

1594 DOI: 10.1021/jf5044025

J. Agric. Food Chem. 2015, 63, 1593−1600
Journal of Agricultural and Food Chemistry Article

Table 2. Reference Standards Used in Descriptive Analysis exclusion criteria precluded participation by wine industry profes-
of Cabernet Sauvignon Wines after Oak Maturation sionals and university staff and students. One hundred and sixteen
consumers were recruited and acceptability tests were performed in
descriptora reference standard October 2012, approximately 3 months after DA was completed.
red fruit two McCain frozen raspberries and one fresh strawberry Consumers were asked to rate their hedonic liking of each wine using a
confectionary 1
/2 Allens strawberries and cream lolly + 0.5 cm piece Coles 9-point scale with anchors from 1 (dislike extremely) to 9 (like
banana lolly extremely). Prior to evaluation of wines, consumers were instructed on
dark fruit two McCain frozen blackberries and two McCain frozen how to assess the wine and use the hedonic scale. Coded samples were
blueberries presented monadically with consumers signaling the researchers for
herbaceous 0.5 cm cube of green capsicum, two small tomato plant their next sample using a switch in the sensory booth. Crackers and
leaves distilled water were provided as palate cleansers. After wines were
mint/menthol 1
/2 tsp eucalyptus solution (1 drop of eucalyptus oil evaluated, consumers answered demographic questions regarding
(Bosisto’s, Australia) in 2 L of water) gender, age, education, household income, and wine consumption.
dustyb 3 cm square of Hessian, 3 sticks of white chalk crushed Consumers also indicated their average spend per bottle of wine for
earthy 1 tsp wet earth with moss home consumption.
charry 1 tsp 4-ethylguaiacol solution (1 drop of 4-ethylguaiacol in The expert panel comprised 10 wine industry personnel (9 males
50 mL of water) and 2 tsp of oak solution (medium- and 1 female), who met the criteria of “expert” defined by Parr and co-
toasted American oak wood (Oak Chips Inc., Waverly, workers.24 Each panelist assessed 10 wines, that is, the subset of 5
OH, USA) prepared in 10% ethanol) wines in duplicate. Participants were asked to assess each wine
leatherb 3 cm × 10 cm piece of leather according to the 20-point scoring system used in Australian wine show
mocha 1 tsp of mocha solution (100 g of 70% Lindt chocolate + 20 judging25 and to provide informal comments describing what they
coffee beans in 250 mL of water) liked or disliked about each wine. Tasting was conducted 4 months
licorice 0.5 cm piece of Darrell Lea licorice after consumer acceptability testing in an open-plan tasting room at a
1
sweet oak /8 tsp of vanillin powder (Sigma-Aldrich) + 1/2 tsp of oak winery in the Barossa Valley.
solution Data Analysis. Data were analyzed using Senpaq v5.01 (Qi
1
spice /4 tsp of McKenzie’s mixed spice + 1/4 tsp of freshly Statistics, 2012) and XLSTAT 2013.1.01 (Addinsoft, 2012).
ground black pepper + 3 cloves Consumer acceptance data were analyzed using a combination of
woodyb 1 cm × 9 cm × 0.5 cm untoasted French oak wood (Oak descriptive and multivariate techniques, including analysis of variance
Chips Inc.) + 1/2 tsp of pencil shavings (ANOVA) with post hoc Tukey’s test, partial least-squares regression
a
Samples prepared in 30 mL of Cabernet Sauvignon cask wine, unless (PLSR), internal preference mapping, and principal component
otherwise stated. bDenotes samples that were prepared dry. analysis (PCA). Hedonic clusters were identified using cluster analysis.

descriptor using a 15 cm unstructured line scale with anchor points of

“low” and “high” placed at 10 and 90% on the scale, respectively.
Sensory reference standards (Table 2) were developed during training
sessions and provided to panelists at each formal assessment session in
covered, opaque black glasses. Panelists were required to reacquaint
themselves with the sensory reference standards and written
definitions of attributes several times during each evaluation. Distilled
water and crackers were provided as palate cleansers, and panelists
were required to have a 1 min break between samples and a 5 min
break every three samples. Sensory data were collected using FIZZ
software (version 2.47b, Biosystemes, Couternon, France).
Consumer Acceptance and Expert Panel Testing. On the basis of
the outcomes of DA, five wines (barrel, SS + OAK, P1 + OAK, P1 −
OAK, and P2VAT), with distinct sensory profiles were selected for
consumer acceptance and expert panel testing. Consumer acceptability
tests were carried out in a sensory laboratory at a central location, with
each participant assigned a random “taster” number and allocated to
an individual booth. Participants were recruited using a variety of
methods, including social networking sites (e.g., Facebook, LinkedIn,
Twitter), distribution of a flyer (via a mailbox drop within a 6 km
radius of the University of Adelaide’s Waite campus, and distribution
at local wine retail outlets) and e-mail. Participants were screened
against inclusion criteria requiring consumption of red wine at least
once a month and being of legal drinking age (i.e., ≥18 years of age);
■ RESULTS AND DISCUSSION
Cabernet Sauvignon wines (a 2009 Padthaway and a 2011 Eden
Valley) were matured using a range of traditional and/or
alternative oak maturation regimens (Table 1). Various
chemical and sensory analyses were subsequently performed
to compare the influence of each maturation regimen on the
composition, sensory properties, quality and consumer accept-
ability of wines.
Influence of Oak Maturation Regimen on Chemical
Composition of Wine. The wine color density, wine hue,
total phenolics, alcohol content, pH, TA and VA of wines are
reported in Table 3. Wines from trial 1 (i.e., 2009 Padthaway
wines) had slightly lower pH, TA and VA and slightly higher
alcohol content than wines from trial 2 (i.e., 2011 Eden Valley
wines). There were no meaningful differences between the
basic wine parameters of trial 1 wines, but trial 2 wines showed
greater variation, in particular for wine color density and
phenolic content, which ranged from 9.7 to 10.2 and from 51 to
57 au, respectively. Increased phenolic content can influence
the perception of astringency by enhancing the drying
sensation,26 but the relatively small differences observed

Table 3. Wine Color Density, Wine Hue, Phenolics, Alcohol Content, pH, Titratable Acidity (TA), and Volatile Acidity (VA) of
Cabernet Sauvignon Wines after Oak Maturation
wine color phenolics alcohol
density wine hue (au) (% v/v) pH TAa (g/L) VAa (g/L)
trial 1, 2009 Padthaway mean 12.6 ± 0.3 0.84 ± 0.01 61 ± 1 14.6 ± 0.01 3.43 ± 0.01 6.1 ± 0.05 0.42 ± 0.02
Cabernet Sauvignon range 12.4−12.8 0.83−0.86 60−62 14.5−14.6 3.43−3.45 6.0−6.2 0.41−0.42

trial 2, 2011 Eden Valley mean 9.9 ± 0.4 0.77 ± 0.01 53 ± 2 14.2 ± 0.05 3.51 ± 0.01 6.4 ± 0.07 0.52 ± 0.15
Cabernet Sauvignon range 9.7−10.2 0.76−0.80 51−57 14.1−14.3 3.50−3.53 6.4−6.6 0.54−0.62

a
TA measured as g/L of tartaric acid; VA measured as g/L of acetic acid.

1595 DOI: 10.1021/jf5044025

J. Agric. Food Chem. 2015, 63, 1593−1600
Journal of Agricultural and Food Chemistry Article

Table 4. Oak Volatile Composition of Cabernet Sauvignon Wines after Oak Maturationa
concentration (μg/L)
cis-oak trans-oak 4-methyl 5-methyl
treatment lactone lactone guaiacol guaiacol eugenol vanillin furfural furfural
trial 1, 2009 Padthaway SSVAT 128 ± 1 a 23 ± 0 7±0b 4±0b 12 ± 1 54 ± 1 b 121 ± 0 nd
Cabernet Sauvignon P1VAT 121 ± 1 b 27 ± 3 8±0a 11 ± 0 a 5±7 153 ± 12 a 123 ± 1 nd
P2VAT 121 ± 1 b 29 ± 0 8±0a 10 ± 0 a nd 128 ± 42 ab 119 ± 3 nd
FOVAT 122 ± 4 b 29 ± 3 8±0a 11 ± 1 a 10 ± 0 131 ± 19 ab 119 ± 2 nd
P value 0.048 0.107 <0.001 <0.001 0.096 0.052 0.243

trial 2, 2011 Eden Valley barrel 99 ± 1 a 79 ± 4 a 4±1c 1±0c 6±9 123 ± 3 b 261 ± 24 c 32 ± 1 d
Cabernet Sauvignon SS + OAK 39 ± 1 c 30 ± 1 c 28 ± 1 b 15 ± 0 b nd 242 ± 5 a 1877 ± 2 b 282 ± 1 b
SS − OAK nd nd 4±1c nd nd 18 ± 7 c 173 ± 1 cd nd
P1 + OAK 30 ± 1 d 46 ± 2 b 27 ± 1 b 16 ± 0 a nd 281 ± 64 a 32 ± 1 e 54 ± 1 c
P1 − OAK nd nd 4±1c nd nd 6±8c 121 ± 4 de nd
P2 + OAK 47 ± 1b 24 ± 2 d 33 ± 1 a 16 ± 0 a 6±8 313 ± 98 a 2633 ± 141 a 368 ± 13 a
P2 − OAK nd nd 3±0c nd nd 20 ± 11 c 116 ± 4 de nd
P value <0.001 <0.001 <0.001 <0.001 0.580 <0.001 <0.001 <0.001
a
Values are the mean of two technical replicates (± standard deviation). Means within a column (per trial) followed by different letters are
significantly different (P = 0.05, one-way ANOVA, Tukey’s LSD post hoc).

Table 5. Mean Intensity Ratings for Sensory Attributes of Cabernet Sauvignon Wines after Oak Maturationa
trial 1 trial 2
2009 Padthaway Cabernet Sauvignon 2011 Eden Valley Cabernet Sauvignon
attributeb SSVAT P1VAT P2VAT FOVAT barrel SS + OAK SS − OAK P1 + OAK P1 − OAK P2 + OAK P2 − OAK
A-red fruit 4.8 5.2 5.5 5.2 7.1 ab 6.8 ab 8.3 a 6.2 b 7.8 a 5.9 b 7.8 a
A-confectionary 4.5 5.3 5.2 4.2 5.7 ab 6.0 ab 6.9 ab 5.5 b 7.0 ab 5.7 ab 7.1 a
A-dark fruit 6.8 7.6 7.2 6.9 6.8 b 6.7 b 7.1 ab 7.3 ab 7.0 ab 6.9 ab 8.2 a
A-herbaceous 3.5 3.3 3.9 3.5 4.1 ab 3.4 b 4.0 ab 4.3 ab 4.9 a 3.2 b 5.2 a
A-mint/menthol 4.1 4.1 4.0 3.6 4.3 bc 4.1 bc 4.4 bc 4.1 bc 5.4 ab 4.0 c 5.9 a
A-dusty 6.7 6.1 6.2 6.7 4.1 abc 4.3 abc 3.5 c 4.9 a 3.7 bc 4.5 ab 4.2 abc
A-charry 4.1 b 4.8 ab 5.1 ab 5.9 a 3.2 b 4.5 a 2.7 b 5.6 a 3.0 b 4.6 a 3.2 b
A-leather 5.2 b 5.0 b 5.4 ab 6.6 a 3.2 ab 3.3 ab 2.5 b 3.6 ab 3.0 ab 3.7 a 3.1 ab
A-mocha 3.5 4.3 4.0 4.2 3.0 c 3.8 bc 3.6 bc 5.0 a 3.5 bc 4.4 ab 3.5 bc
A-sweet oak 4.4 5.1 4.9 4.7 4.8 ab 5.1 ab 4.6 b 5.8 a 5.1 ab 5.7 ab 4.9 ab
A-woody 5.1 5.9 6.4 5.8 4.0 b 4.9 ab 3.9 b 5.4 a 4.0 b 5.2 ab 4.2 ab

P-red fruit 6.5 6.0 6.2 7.2 8.1 ab 7.0 ab 8.2 ab 6.9 b 7.9 ab 7.0 b 8.3 a
P-earthy 3.3 3.8 3.5 3.6 2.8 b 3.5 a 2.8 ab 3.3 ab 3.0 ab 3.2 ab 3.4 ab
P-charry 3.3 b 5.2 a 4.9 a 5.0 a 3.7 bc 4.4 ab 3.0 c 5.0 a 3.3 c 4.6 ab 3.7 bc
P-mocha 3.5 4.3 4.2 4.3 3.8 ab 4.2 ab 3.7 ab 4.4 a 3.4 b 4.3 a 3.4 b
P-licorice 3.7 3.9 3.2 3.6 3.5 bcd 2.9 d 4.4 a 3.9 abc 4.3 ab 3.5 cd 3.9 abc
P-sweet oak 4.7 ab 5.4 a 4.4 b 4.9 ab 4.7 bc 4.7 bc 4.2 c 5.6 ab 4.4 c 6.5 a 4.9 bc
P-woody 5.6 b 7.4 a 6.6 ab 6.4 ab 4.8 b 6.7 a 5.3 ab 5.9 ab 5.5 ab 5.9 ab 5.1 ab
tannins 7.9 b 8.5 ab 9.4 a 9.0 ab 7.2 ab 7.3 a 6.7 ab 7.2 ab 7.4 a 6.0 b 7.3 ab
a
Values are mean scores from one wine replicate (per treatment) presented to 10 judges in three replicate sessions. Means within a row (per trial)
followed by different letters are significantly different (P = 0.05, one way ANOVA, Tukey’s LSD post hoc). bA, aroma attribute; P, palate attribute.

between phenolic measurements obtained in the current study
(i.e., 5−10%) would not be expected to lead to perceptible
sensory differences.
Similar concentrations of oak volatiles were observed in each
of the trial 1 wines, albeit SSVAT wines contained significantly
lower levels of vanillin than the other wines from trial 1 (Table
4). Of the oak volatiles measured, only cis-oak lactone and
eugenol occurred at concentrations above their detection
thresholds, of 23 and 5 μg/L, respectively.27,28 More significant
differences were observed between trial 2 wines. Wines aged
using maturation regimens without the addition of oak wood
(i.e., SS − OAK, P1 − OAK, and P2 − OAK) contained low or
undetectable (<1 μg/L) levels of oak volatiles, as expected. The
1596
barrel-aged wine contained the highest concentrations of cis-
and trans-oak lactones (99 and 79 μg/L respectively), but
comparatively lower levels of other oak-derived volatiles. The
maturation regimens involving addition of oak staves yielded
similar oak volatile concentrations, albeit P1 + OAK wines
contained extremely low levels of furfural and 5-methylfurfural
(i.e., <100 μg/L) compared to SS + OAK and P2 + OAK wines
(which contained >1800 μg/L) and even the unoaked wines.
Again, cis-oak lactone was detected (in oaked wines) at
concentrations above its detection threshold. The guaiacol and
vanillin levels of trial 2 wines aged using alternative oak
maturation regimens also exceeded threshold concentrations,
being 10 and 200 μg/L, respectively.28
DOI: 10.1021/jf5044025
J. Agric. Food Chem. 2015, 63, 1593−1600
Journal of Agricultural and Food Chemistry
Figure 1. PCA biplot of oak volatile concentrations and sensory attribute ratings for Cabernet Sauvignon wines after oak maturation.
The differences observed between the oak volatile profiles of
wines from trial 1 compared to those from trial 2 likely reflect
differences in oak composition, in agreement with previous
studies.29−31 Interestingly, the oak wood used in trial 1 yielded
a considerably higher proportion of cis-oak lactone (i.e.,
approximately 4-fold higher levels) than trans-oak lactone,
whereas the oak used in trial 2 gave approximately equal
concentrations of each isomer. Compositional differences can
also be attributed to differences in the rate of oak addition, the
surface area available for extraction, and the duration of oak
maturation. Surprisingly, meaningful differences in wine color,
phenolic and oak volatile profiles were not observed for the
different maturation regimens, even when oxygen ingress varied
considerably (Table 1).
Influence of Oak Maturation Regimen on Sensory
Properties of Wine. ANOVA was performed on the DA data,
and 19 of 31 attributes were found to significantly differentiate
(P < 0.05) wines within each trial (Table 5).
Wines from trial 1 were generally rated higher in dusty,
charry, leather, and woody characters than trial 2 wines, with
FOVAT receiving the highest ratings for charry and leather
aromas. Similar ratings were given to varietal attributes, that is,
red f ruit, dark f ruit, and herbaceous attributes. That wines from
this trial were perceived to be more complex, that is, to exhibit
more intense oak- and age-related sensory attributes, reflects
the extended period of oak maturation (12 months) and bottle
aging (24 months) prior to sensory analysis. The duration of
oak maturation directly influences the concentration of volatiles
extracted from oak wood into wine,32 but biochemical reactions
continue to occur, for example, the conversion of furanic
aldehydes into their corresponding alcohols.33,34 This likely
explains the lower levels of furfural (i.e., approximately 120 μg/
L) and the absence of 5-methylfurfural observed in trial 1
wines, compared to trial 2 wines, in particular SS + OAK and
P2 + OAK (Table 4).
The unoaked wines from trial 2 received similar ratings for
most attributes and tended to exhibit more intense varietal
characteristics, for example, red f ruit, dark f ruit, herbaceous, and
mint/menthol attributes, than the oaked wines; the prominence
of varietal characters is due to the absence of oak influence. Of
the oak-treated wines, the barrel-aged wine received the lowest
scores for most oak-related attributes, which was not surprising
given the low levels of oak volatiles observed in this wine. The
1597
Article
ratings for oak attributes in the barrel wine more closely
resembled scores given to the unoaked wines, and instead, the
barrel wine displayed more intense red f ruit and mint/menthol
aromas and red f ruit on the palate. P1 + OAK exhibited the
highest charry, mocha, sweet oak, and woody aromas and charry,
licorice flavor, whereas P2 + OAK displayed leather and sweet
oak on the palate. Previous studies have investigated the impact
of micro-oxygenation (MOX) on wine sensory attributes,
including the development of aged notes in MOX wines. For
example, Cejudo-Bastante and colleagues observed an increase
in nutty, tobacco, and spice notes in MOX-treated wines
compared to wines stored in stainless steel tanks without
oxygen addition.35 Several of the maturation regimens used in
trial 2 deliberately enabled oxygen ingress (Table 1), but the
resulting wines did not exhibit overly apparent aged notes.
PCA was performed on the mean intensity ratings obtained
for the sensory attributes that were perceived to be significantly
different across the two trials (Figure 1), with PC1 and PC2
accounting for 44 and 22% of variation between wines,
respectively. Clear separation was observed between oaked
wines, which were predominantly located in the right
quadrants, and unoaked wines, which were located in the left
quadrants. The barrel-aged wine was located in close proximity
to the unoaked wines, which again was not surprising given its
relatively low levels of oak volatiles (Table 4). PC1 suggests
separation of oaked and unoaked wines is primarily driven by
the intensity of varietal (i.e., f ruit/green) versus oak-related
sensory attributes. Discrimination of trial 1 and 2 wines was
also observed, with wines aged in vats clustered together in the
bottom right quadrant and wines aged in stainless steel or
plastic tanks (containing oak) clustered in the top right
quadrant. Trial 1 wines were in part differentiated on the basis
of dusty and leather aromas, which reflect the increased age of
these wines. SS + OAK, P1 + OAK, and P2 + OAK each
contained higher levels of vanillin (being 242−313 μg/L) than
trial 1 wines (Table 4). These concentrations exceed the
detection threshold reported for vanillin in red wine, that is,
200 μg/L.28 Vanillin might therefore contribute to the sweet oak
and mocha attributes that helped differentiate these wines.
Differentiation of trial 1 and 2 wines may also have been
influenced by guaiacol, cis-oak lactone, and eugenol, that is, the
other volatiles observed in selected wines at concentrations
above their threshold levels (Table 4).
DOI: 10.1021/jf5044025
J. Agric. Food Chem. 2015, 63, 1593−1600
Journal of Agricultural and Food Chemistry Article

Influence of Oak Maturation Regimen on Quality and Table 7. Consumer Demographics for Total Sample (n =
Consumer Acceptance of Wines. A subset of five wines was 116) and Consumer Segments
selected for quality and acceptability ratings by wine experts
total samplea cluster 1a cluster 2a cluster 3a
and consumers, respectively. The expert panel used a 20-point (n = 116) (n = 35) (n = 55) (n = 26)
scoring system to rate wine quality (Table 6); mean scores
gender
male 50.9 42.9 63.6 34.6
Table 6. Consumer Liking and Expert Quality Ratings for female 49.1 57.1 36.4 65.4
Selected Cabernet Sauvignon Wines after Oak Maturationa age (years)
quality 18−34 27.6 37.1 20.0 30.8
hedonic ratingsb ratingsb 35−54 31.9 28.6 27.3 46.2
total wine 55+ 40.5 34.3 52.7 23.1
sample cluster 1 cluster 2 cluster 3 experts education
treatment (n = 116) (n = 35) (n = 55) (n = 26) (n = 10)
high school 10.3 17.1 7.3 7.7
barrel 5.7 5.11 b 6.35 a 4.85 b 13.75 technical/trade 15.5 5.7 21.8 15.4
SS + OAK 5.8 4.49 b 6.45 a 6.12 a 14.42 bachelor’s degree 29.3 28.6 30.9 26.9
P1 − OAK 5.9 4.31 b 6.60 a 6.23 a 14.82 postgraduate 44.8 48.6 40.0 50.0
P1 + OAK 5.9 3.91 b 6.73 a 6.58 a 13.92 household income ($AUD)
P2VAT 5.9 5.80 b 7.00 a 3.42 c 14.65 <$25000 8.6 11.4 7.3 7.7
a
Means within a row followed by different letters are significantly $25001−50000 11.2 5.7 14.5 11.5
different (P = 0.05, one-way ANOVA, Tukey’s LSD post hoc). $50001−75000 18.1 25.7 16.4 11.5
b
Hedonic ratings were determined using a 9-point scale and quality $75001−100000 21.6 25.7 18.2 23.1
ratings using a 20-point scoring system. $100001−150000 28.4 20.0 32.7 30.8
$150001−200000 7.8 5.7 7.3 11.5
ranged from 13.8 (for the barrel-aged wines) to 14.8 (for P1 − >$200000+ 4.3 5.7 3.6 3.8
OAK), which indicates all wines were technically sound. The wine consumption
experts informally described the unoaked sample as having 4+ times/week 22.2 22.9 27.3 11.5
“varietal characters, f resh cherry nose, forest f ruits, confectionary” 2−3 times/week 35.9 28.6 40.0 38.5
but “lacking complexity and palate length”. The barrel-aged wine once a week 27.4 28.6 23.6 34.6
was considered to display “dull, manufactured oak” and “lack once every 2 weeks 11.1 17.1 7.3 11.5
f ruit intensity”. P2VAT was described as “savory, chocolate, dusty, once a month 2.6 2.9 1.8 3.8
earthy, lacking f ruit f lavor” with “sof ter tannins” and “more average price/bottle for home consumption ($AUD)
complex than the other wines”. In general, these comments were <$10 2.6 2.9 1.8 3.8
in good agreement with DA results. $11−15 26.5 25.7 27.3 26.9
One hundred and sixteen consumers participated in the $16−20 37.6 40.0 41.8 26.9
acceptance test, and their demographic data are reported in $21−30 28.2 28.6 25.5 34.6
Table 7. A similar proportion of male and female participants $31−50 3.4 2.9 1.8 7.7
and a relatively even distribution of ages were achieved. Most $51+ 0.9 0 1.8 0
consumers were regular wine drinkers and consumed wine at a
Values expressed as percentages.
least once per week (i.e., 86%). There were no significant
differences in the hedonic ratings given by the consumer cohort
for the five wines (Table 6), with mean scores ranging from 5.7 drivers were observed for cluster 1, including leather aroma
to 5.9. However, as in previous research,17 disparity in (0.34) and flavor (0.33), licorice flavor (0.24), and astringent/dry
consumers’ wine preferences was observed, and cluster analysis mouthfeel (0.26), whereas strong negative drivers included
based on individual hedonic scores enabled the identification of mocha (−0.47) and charry (−0.30) aromas. This explains
three distinct consumer segments (Tables 6 and 7). The first cluster 1’s preference for P2VAT, which exhibited more
cluster comprised 35 consumers, who were not overly accepting apparent aged characters relative to other wines. Positive
of any wine, but who liked P2VAT most and P1 + OAK least. drivers for cluster 3 included red f ruit (0.42), charry (0.27), and
Cluster 2, the largest cluster (n = 55), gave similar scores to mocha (0.56) aromas and earthy (0.61) and woody (0.31)
each wine and quite closely resembled the total sample. Cluster flavors. The strong negative drivers for cluster 3 were dusty
3 comprised 26 consumers, who did not like the barrel-aged or (−0.44) and leather (−0.64) aromas and leather flavor on the
P2VAT wines, but equally liked the other wines. Surprisingly, palate (−0.57). Consumers within this cluster tended to be
the wine produced using traditional barrel maturation was not younger, from households with higher incomes, who were
favored by any segment and actually received the lowest overall willing to spend more for wine on average than consumers
score, whereas P1 − OAK, the wine made without oak contact, from clusters 1 and 2 (Table 7). On the basis of demographic
was quite well liked by clusters 2 and 3. data, hedonic scores, and the sensory attributes driving wine
Internal preference mapping was undertaken to investigate preferences, consumers from clusters 2 and 3 would be most
the sensory attributes driving consumer liking. PLSR was likely to be receptive to wines made using alternate oak
performed on the DA data and hedonic ratings for each maturation regimens; indeed, consumers from cluster 2 were
consumer segment and the total sample (Figure 2). Sensory generally accepting of all wines (Table 6).
attributes with regression coefficients > ±0.2 are generally This study demonstrates consumer acceptance of wines
considered to influence consumer liking.6 In the current study, made using alternative oak maturation regimes. The presence
there were neither positive nor negative drivers for either the of key oak-derived volatile compounds and perceptible oak
total consumer sample or cluster 2. In contrast, strong positive aromas and flavor in oak-aged wines was confirmed by chemical
1598 DOI: 10.1021/jf5044025
J. Agric. Food Chem. 2015, 63, 1593−1600
Journal of Agricultural and Food Chemistry
Figure 2. PLSR plot of consumer liking against wine sensory attributes for selected Cabernet Sauvignon wines.
and sensory analysis. Differences observed in the oak volatile
concentrations and intensity of varietal versus oak-derived
aromas and flavors of wines reflect differences in oak treatment,
including the duration of oak maturation and the use of
different oak woods. Quality rating scores ranged from 13.8 to
14.8, indicating wines were considered technically sound, but
because wines aged using oak alternatives tend not to be
premium quality wines, quality scores were always expected to
be modest. Segmentation based on consumers’ hedonic ratings
demonstrated the variation in wine preferences of different
consumers, but importantly, also demonstrated the appeal of
wines made using alternative maturation regimens to
consumers within different segments. These results justify
wine producers’ use of alternative oak maturation regimens, in
particular for the production of wines targeted at certain price
points and/or toward specific segments of the wine market.
Furthermore, this study highlights the importance of combining
compositional and descriptive sensory analyses with consumer
research (e.g., hedonic ratings) to achieve a more substantial
overview of the drivers of wine quality and acceptability.
■
AUTHOR INFORMATION
Corresponding Author
*(K.L.W.) Fax: + 61 8 8313 7116. E-mail: kerry.wilkinson@
adelaide.edu.au.
Funding
A.M.C. thanks the Australian Grape and Wine Authority for
provision of a research scholarship.
Notes
The authors declare no competing financial interest.
■ ACKNOWLEDGMENTS
We thank the producers who kindly donated the wines used in
this study; without their support this research would not have
been possible. We acknowledge the technical support provided
by industry partners and the panelists who participated in the
DA, expert, and consumer tastings.
■
REFERENCES
(1) Mueller, S.; Szolnoki, G. The relative influence of packaging,
labelling, branding and sensory attributes on liking and purchase
intent: consumers differ in their responsiveness. Food Qual. Pref. 2010,
21, 774−783.
1599
Article
(2) Marin, A. B.; Jorgensen, E. M.; Kennedy, J. A.; Ferrier, J. Effects
of bottle closure type on consumer perceptions of wine quality. Am. J.
Enol. Vitic. 2007, 58, 182−191.
(3) Johnson, T. E.; Bastian, S. E. P. A preliminary study of the
relationship between Australian wine consumers’ expertise and their
wine purchasing and consumption behaviour. Aust. J. Grape Wine Res.
2007, 13, 186−197.
(4) Chrea, C.; Melo, L.; Evans, G.; Forde, C.; Delahunty, C.; Cox, D.
N. An investigation using three approaches to understand the
influence of extrinsic product cues on consumer behaviour: an
example of Australian wines. J. Sens. Stud. 2011, 26, 13−24.
(5) Bruwer, J.; Li, E. Wine-related lifestyle (WRL) market
segmentation: demographic and behavioural factors. J. Wine Res.
2007, 18, 19−34.
(6) Lattey, K. A.; Bramley, B. R.; Francis, I. L. Consumer
acceptability, sensory properties and expert quality judgements of
Australian Cabernet Sauvignon and Shiraz wines. Aust. J. Grape Wine
Res. 2010, 16, 189−202.
(7) Mueller, S.; Lockshin, L.; Saltman, Y.; Blanford, J. Message on a
bottle: the relative importance of wine back label information on wine
choice. Food Qual. Pref. 2010, 21, 22−32.
(8) Arfelli, G.; Sartini, E.; Corzani, C.; Fabiani, A.; Natali, N. Impact
of wooden barrel storage on the volatile composition and sensorial
profile of red wine. Food Sci. Technol. Int. 2007, 13, 293−299.
(9) Chatonnet, P.; Boidron, J. N.; Pons, M. Maturation of red wines
in oak barrels: evolution of some volatile compounds and their
aromatic impact. Sci. Aliments 1990, 10, 565−587.
(10) Ortega-Heras, M.; Rivero-Perez, M. D.; Pérez-Magariño, S.;
González-Huerta, C.; González-Sanjosé, M. L. Changes in the volatile
composition of red wines during aging in oak barrels due to
microoxygenation treatment applied before malolactic fermentation.
Eur. Food Res. Technol. 2008, 226, 1485−1493.
(11) Ortega-Heras, M.; González-Huerta, C.; Herrera, P.; González-
Sanjosé, M. L. Changes in wine volatile compounds of varietal wines
during ageing in wood barrels. Anal. Chim. Acta 2004, 513, 341−350.
(12) Lockshin, L. S.; Rhodus, W. T. The effect of price and oak flavor
on perceived wine quality. Int. J. Wine Market 1993, 5, 13−25.
(13) Cano-López, M.; Bautista-Ortín, A. B.; Pardo Mínguez, F.;
López-Roca, J. M.; Gómez-Plaza, E. Sensory descriptive analysis of a
red wine aged with oak chips in stainless steel tanks or used barrels:
effect of the contact time and size of the oak chips. J. Food Qual. 2008,
31, 645−660.
(14) Gutiérrez-Afonso, V. L. Sensory descriptive analysis between
white wines fermented with oak chips and in barrels. J. Food Sci. 2002,
67, 2415−2419.
(15) Bozalongo, R.; Carillo, J. D.; Torroba, M. A. F.; Tena, M. T.
Analysis of French and American oak chips with different toasting
DOI: 10.1021/jf5044025
J. Agric. Food Chem. 2015, 63, 1593−1600
Journal of Agricultural and Food Chemistry Article

degrees by headspace solid-phase microextraction-gas chromatogra-

phy-mass spectrometry. J. Chromatogr., A 2007, 1173, 10−17.
(16) Arfelli, G.; Sartini, E.; Corzani, C.; Fabiani, A. Chips, lees, and
micro-oxygenation: influence on some flavours and sensory profile of a
bottles red Sangiovese wine. Eur. Food Res. Technol. 2011, 233, 1−10.
(17) Pérez-Magariño, S.; Ortega-Heras, M.; González-Sanjosé, M. L.
Wine consumption habits and consumer preferences between wines
aged in barrels or with chips. J. Sci. Food Agric. 2011, 91, 943−994.
(18) Crump, A. M.; Johnson, T. E.; Bastian, S. E. P.; Bruwer, J.;
Wilkinson, K. L. Consumers’ knowledge of and attitudes towards the
role of oak in winemaking. Int. J. Wine Res. 2014, 6, 21−30.
(19) Iland, P.; Bruer, N.; Edwards, G.; Weeks, S.; Wilkes, E. Chemical
Analysis of Grapes and Wine: Techniques and Concepts; Patrick Iland
Wine Promotions Pty. Ltd.: Campbelltown, Australia, 2004.
(20) Pollnitz, A. P.; Pardon, K. H.; Sefton, M. A. Quantitative analysis
of 4-ethylphenol and 4-ethylguaiacol in red wine. J. Chromatogr., A
2000, 874, 101−109.
(21) Pollnitz, A. P.; Pardon, K. H.; Sykes, M.; Sefton, M. A. The
effects of sample preparation and gas chromatograph injection
techniques on the accuracy of measuring guaiacol, 4-methylguaiacol
and other volatile oak compounds in oak extracts by stable isotope
dilution analyses. J. Agric. Food Chem. 2004, 52, 3244−3252.
(22) Pollnitz, A. P. The Analysis of Volatile Wine Components Derived
from Oak Products during Winemaking and Storage; The University of
Adelaide, Australia, 2000.
(23) Lawless, H. T.; Heymann, H. Acceptance and preferance testing.
In Sensory Evaluation of Food: Principles and Practices; Heldman, D. R.,
Ed.; Chapman & Hall, International Thomson Publishing: New York,
1999; pp 430−479.
(24) Parr, W. V.; Heatherbell, D.; White, K. G. Demystifying wine
expertise: olfactory threshold, perceptual skill and semantic memory in
expert and novice wine judges. Chem. Sens. 2002, 27, 747−755.
(25) Gawel, R.; Godden, P. W. Evaluation of the consistency of wine
quality assessments from expert wine tasters. Aust. J. Grape Wine Res.
2008, 14, 1−8.
(26) Gawel, R. Red wine astringency: a review. Aust. J. Grape Wine
Res. 1998, 4, 74−95.
(27) Wilkinson, K. L.; Elsey, G. M.; Prager, R. H.; Tanaka, T.; Sefton,
M. A. Precursors to oak lactone. Part 2: Synthesis, separation and
cleavage of several β-D-glucopyranosides. Tetrahedron 2004, 60, 6091−
6100.
(28) Guth, H. Quantitation and sensory studied of character impact
odorants of different white wine varieties. J. Agric. Food Chem. 1997,
45, 3027−3032.
(29) Spillman, P. J.; Iland, P. G.; Sefton, M. A. Accumulation of
volatile oak compounds in a model wine stored in American and
Limousin oak barrels. Aust. J. Grape Wine Res. 1998, 4, 67−73.
(30) Pollnitz, A. P.; Jones, G. P.; Sefton, M. A. Determination of oak
lactones in barrel aged wines and in oak extracts by stable isotope
dilution analysis. J. Chromatogr., A 1999, 857, 239−246.
(31) Mosedale, J. R.; Puech, J. L.; Feuillat, F. The influence on wine
flavour of the oak species and natural variation of heartwood
components. Am. J. Enol. Vitic. 1999, 50, 503−512.
(32) Garde-Cerdán, T.; Ancín-Azpilicueta, C. Effect of oak barrel
type on the volatile composition of wine: storage time optimization.
LWT−Food Sci. Technol. 2006, 39, 199−205.
(33) Spillman, P. J.; Pollnitz, A. P.; Liacopoulos, D.; Pardon, K. H.;
Sefton, M. A. Formation and degradation of furfuryl alcohol, 5-
methylfurfuryl alcohol, vanillyl alcohol, and their ethyl ethers in barrel
aged wines. J. Agric. Food Chem. 1998, 46, 657−663.
(34) Garde-Cerdán, T.; Ancín-Azpilicueta, C. Review of quality
factors on wine ageing in oak barrels. Trends Food Sci. Technol. 2006,
17, 438−447.
(35) Cejudo-Bastante, M. J.; Hermosín-Gutiérrez, I.; Peréz-Coello,
M. S. Micro-oxygenation and oak chip treatments of red wines: effects
on colour-related phenolics, volatile composition and sensory
characteristics. Part II: Merlot wines. Food Chem. 2011, 124, 738−748.

1600 DOI: 10.1021/jf5044025

J. Agric. Food Chem. 2015, 63, 1593−1600